-
Notifications
You must be signed in to change notification settings - Fork 12
Open
Labels
Description
Category
Dashboard visualization
Description
After doing a POC with KQL queries from Azure Monitor, we should now try and create a production-ready anomaly detection system in Invictus.
The obective is to be able to detect possible spikes in
-
Traffic: detect an unusual amount of messages passing through the Integrations as compared to normal operation
-
Errors: detect an unusual amount of errors occurring
-
IDEA: Provide a visualisation (graph?) in the Dashboard itself
-
IDEA: Hook into the existing Invictus Alerts, so that a user can get alerted when an anomaly is detected
Additional context
No response